Because AI assistants often send no referrer header, you cannot trace an AI-driven enquiry in Google Analytics. The audit solves this by measuring share of AI recommendations before and after, and by tying visibility changes to new patient or client wins.
TL;DR
The referrer problem
When a prospect asks an AI assistant a question and clicks through to your website, the request often arrives with an empty referrer. This is not a tracking failure — it is how AI assistants work. They do not pass along the source of the answer because the answer is generated, not a crawled link. In analytics, these visits look like direct traffic, making it impossible to say “this enquiry came from an AI answer.”
Why traditional metrics fail
- Direct traffic is not actionable. You cannot tell whether a direct visit came from an AI answer, a bookmark, or a typed URL.
- No conversion path. Even if you see a spike in direct traffic, you cannot attribute it to AI visibility work.
- Vanity metrics. Impressions, reach, and “AI mentions” are not tied to business outcomes unless you measure the shortlist and the enquiry.
How the audit measures the business impact
- Share of AI recommendations — the percentage of times your practice is named in the assistant’s answer for a set of prospect questions. Measured before and after remediation.
- Before/after enquiry tracking — the audit captures the questions your prospects actually ask, so you can run the same set yourself and see whether you appear.
- Closed‑loop reporting — when a new patient or client signs up, you can ask which question they asked the AI assistant. If it matches the audit set, you have a direct line from visibility work to revenue.
What the audit delivers
- Monthly share-of-voice report: your share of AI recommendations on each surface, before and after, with screenshots.
- Prompt set: the exact questions your prospects ask, in English and 中文, logged against each assistant.
- Error list: every factual error the assistants state about you, with fixes.
- Optional: a simple spreadsheet to log new enquiries and the question that produced them.
How to use the data
- Plot your share of AI recommendations month over month. An upward trend means more prospects see your name in the AI answer.
- Correlate spikes in new patients with months when your share increased.
- Use the prompt set to test changes yourself: add a directory listing, wait for the next crawl, and re‑run the questions to see if you appear.
SimplyAI (2026). Measuring AI Visibility Against Revenue with No Referrer. Updated 18 September 2026. https://simplyai.work/learn/measuring-ai-visibility-against-revenue-no-referrer/